Artificial Intelligence Brings Resilience and Affordability to the Grid

Share Button

A new white paper from Veritone presents five artificial intelligence (AI)-powered solutions that help those in the electric power industry enhance grid resilience, increase the rate of decarbonization and improve affordability of delivered electricity.

Veritone

Download the full report.

The electric power industry currently faces three key challenges: the need for increased resilience in the face of natural disasters such as wildfires and extreme weather, increased pressure to reduce its carbon output, and cost containment. In its report, Veritone suggests that, when applied to grid management, AI could help address all of these challenges for utilities, independent power producers and developers. The end result would be a cleaner, more reliable and efficient grid.

Veritone explains the benefits that AI can provide to a number of aspects of energy management and generation. The first key Veritone looks at is distributed renewable generation. While distributed renewable generation has transformed the industry, it does present a unique set of new challenges. “The amount of data that needs to be processed in real time to effectively address the intermittency of distributed renewables is overwhelming. No human being could do it,” according to Veritone. But artificial intelligence offers real-time data processing, the ability to generate “predictive models of weather, electricity supply and demand, and pricing,” real-time control and synchronization of both the grid and end-user devices.

AI solutions can also help with distributed energy storage by optimizing the control of individual battery energy systems. This is the second key described in the paper.

“Using real-time load and market data, an AI-based intelligent controller can optimize battery performance, maintain battery health, extend battery life and optimize battery/inverter system operation. An additional benefit to the main grid (and the battery owner) is maximized ancillary services revenue where ancillary services markets exist, including VAR control, frequency control and others.” — Veritone, “Five Keys to Managing the Power Grid with AI

According to Veritone, the third key involves AI solutions that can help with vehicle-to-grid services such as voltage regulation. Electric vehicles (EVs) represent a largely untapped resource for grid resilience as they can be an intermittent storage resource. The fourth key presented in the white paper is that AI can optimize demand response. The final key Veritone presents is how AI can provide “real-time synchronization and control of microgrid resources, including PV, DES, EVs and other DER.” By integrating AI with the power grid, the entire energy ecosystem can benefit, according to the authors.

The use of AI is illustrated in the white paper in several mini case studies. They look at how AI can benefit islanded microgrids, virtual power plants, building smart sensors and electric vehicles.

Download the full report, “Five Keys to Managing the Power Grid with AI,” to learn more about how artificial intelligence can connect green generation and storage assets to create an interconnected, on demand, secure and autonomous electrical grid.

Share Button

Comments

  1. “AI solutions can also help with distributed energy storage by optimizing the control of individual battery energy systems. This is the second key described in the paper.”
    The ‘term’ AI is used a little Cavalierly these days. What is in place now are professional systems with bottom up programming using algorithms and data tables to ‘make’ decisions at computer clock speeds from multi-processor machines. When one gets into (real) AI one gets into the realm of decisions that have had good outcomes and experiences that have had, bad outcomes. The AI taking this “collective” knowledge will form it’s own reactive or proactive protocols, based on these experiences. Bias will creep in at some point and somehow, this has to be kept in check for overall better results. Outside forces like rules, regulations and ordinances may have to be considered to keep the system out of legal trouble. Something like selling excess electricity to another entity when there’s “over the fence” rules in place that determine who and where one can sell electricity to. Now if I was a CFO of a large electric utility, I’d sure want to have a bias in the AI system that gathered the most money per kWh sold and whether the AI rounded up from the thousandth of a cent to the nearest dollar, it would be biased towards the utility’s bottom line not the ratepayer.

    I’ m skeptical, with that kind of computing and control power using a ‘collective’ as variable data points, a simple ‘denial of service’ attack could give the AI system a mental breakdown. Now that’s something one doesn’t want in their massive control system of the electric grid. Now every once in a while the software gets a disk check and compression and clean up to allow more disk space, in the future, what do you say, just forget it and move on?